As much of world is practicing “social distancing” or some version of it, I would venture to guess that many are getting “cabin fever” from the extended isolation. For me this condition has led to a blurred line between work and home and more interestingly news coverage and formal applications.
You are verifying a complex AI or networking chip and found a test failing due to transaction or packet mismatch by scoreboards. As a verification engineer, you would celebrate that you broke the core design intent and found a bug! After hours/days of debugging, all that’s found is a signal on AHB/AXI interface was not connected or a protocol was not followed correctly. Not really a highly effective use of everyone’s time, is it?
Just over a year ago I wrote a blog about the impact of machine learning (ML) algorithms to boost Formal Verification performance . The data for that blog was firsthand experience on a set of complex benchmarks. The data was amazing and quite convincing, but when I wrote that blog the reader needed to “trust” me as the data could not be shared publicly.
Wouldn’t it be nice to get a head start on some things in life? How great would it be to just be able to walk straight to the front of any queue you find yourself in? For me, I’d like a head start on those long flights from the UK to California. If I could start them somewhere over the Rocky Mountains, then it would be a much more pleasant journey… aside from the mountain waves turbulence! While money, fame or just downright rudeness can potentially get you to the front of a queue, I’m going to have to wait for someone to invent teleportation to cut down that journey time.
You may be concerned about the widely published Spectre and Meltdown vulnerabilities affecting most processors, and if your phone and computer are OK. Or more importantly, if you are designing or verifying SoCs, do you have a specter in your design? Let’s first look at what these two vulnerabilities are and how they are affecting your system.
Imagine the scene. It’s Friday night, and you’ve decided to relax and watch a movie. Given the overwhelming amount of choices, you’ve already spent over an hour watching trailers to choose the movie and you’re finally almost ready to go. All that’s left is the popcorn. You go over to the microwave and get it going. For a little while nothing happens, but then you start to hear the pops, slowly at first but then much more rapidly before beginning to taper off. You then have to ask yourself: When do I stop? This is a big question. Take it out too soon and you’re going to break your teeth on those unpopped kernels. Leave it in too long and you risk burning it. How can you know if its popped long enough?
As we have discussed in several of the blogs on this forum, successful deployment of Formal verification requires knowing where and how to use it. Building up an arsenal of techniques that can be applied to deal with complexity and knowing how to use them safely is a necessity for every expert Formal engineer.
Cooking can be a necessity, hobby or calming therapy depending on whom you talk to. Personally speaking, I cook occasionally but even when I am not cooking and I am just a mere silent admirer of this amazing process, onion peeling/cutting/chopping brings tears to my eyes 😀 There are few tricks that one could use to avoid or minimize tears while peeling onions. Some of these tricks work, some don’t.
Someday, in the not too distant future, I will be able to fall asleep, play computer games or write a bestselling novel at the wheel (well 2/3 isn’t bad). Until such time however, I have just the one option – concentrate deeply and blast the speakers with my classic rock and punk collection.
Posted in Property Verification
When running Formal Property Verification, we often see goals that are neither proven nor failing (especially on complex properties), which implies inconclusive goals, also referred to as bounded proofs. In these scenarios, what we have at hand is the Formal bounded depth (in terms of clock cycles), associated with such inconclusive properties.